Locational (In-)Efficiency of Renewable Power Generation Feeding in the Electricity Grid: A Spatial Regression Analysis

This paper analyzes the negative external effects caused by the introduction of variable renewable energy sources into an inflexible power system. We investigate the costs that arise due to the need for temporarily reducing their output to relief grid overstress in the case of high electricity feed-in. The responsible system operator has to remunerate the power plant operator for this lost output. The resulting costs for the system operator, the so-called feed-in management costs, are passed on to the end-consumers in the respective region via the grid use tariff scheme. In this paper, we develop a two-part regression model that explains (i) the occurrence of feed-in management and (ii) the regional variation of feed-in management costs. In the first part, we use a logit model to explain why some regions experienced feed-in management in recent years and others did not. The second part covers an augmented spatial econometric model that investigates the interregional variability of feed-in management costs. The estimates of both models show that especially the installed capacity of wind energy connected to the medium and high voltage level have a negative impact on feed-in management and that high load in a region reduces the need for feed-in management measures. The augmented spatial model indicates for the case of four major DSOs in Germany that an increase by 1 MW of wind energy capacity at the medium and high voltage level lead to an increase in feed-in management costs by 1.9% and 0.9% in regions that already experienced feed-in management, respectively. The policy implication of this finding is that price signals for the reinforcement of the grid infrastructure or for the siting of VRES should be given.